This is an electronic reprint of the original article.This reprint may differ from the original in pagination and typographic detail.

Powered by TCPDF (www.tcpdf.org)

This material is protected by copyright and other intellectual property rights, and duplication or sale of all or part of any of the repository collections is not permitted, except that material may be duplicated by you for your research use or educational purposes in electronic or print form. You must obtain permission for any other use. Electronic or print copies may not be offered, whether for sale or otherwise to anyone who is not an authorised user.

Schulz, F.; Drost, R.; Hämäläinen, Sampsa; Demonchaux, T.; Seitsonen, A.P.; Liljeroth, P.Epitaxial hexagonal boron nitride on Ir(111): A work function template

Published in:Physical Review B

DOI:10.1103/PhysRevB.89.235429

Published: 01/01/2014

Document VersionPublisher's PDF, also known as Version of record

Please cite the original version:Schulz, F., Drost, R., Hämäläinen, S., Demonchaux, T., Seitsonen, A. P., & Liljeroth, P. (2014). Epitaxialhexagonal boron nitride on Ir(111): A work function template. Physical Review B, 89, 1-8. [235429].https://doi.org/10.1103/PhysRevB.89.235429

https://doi.org/10.1103/PhysRevB.89.235429https://doi.org/10.1103/PhysRevB.89.235429

PHYSICAL REVIEW B 89, 235429 (2014)

Epitaxial hexagonal boron nitride on Ir(111): A work function template

Fabian Schulz,1 Robert Drost,1 Sampsa K. Hämäläinen,1 Thomas Demonchaux,1,2 Ari P. Seitsonen,3 and Peter Liljeroth1,*1Department of Applied Physics, Aalto University School of Science, P.O.Box 15100, 00076 Aalto, Finland

2Institut d’Electronique et de Microélectronique et de Nanotechnologies, IEMN, CNRS, UMR 8520, Département ISEN, 41 bd Vauban,59046 Lille Cedex, France

3Institut für Chemie, University of Zürich, Winterthurerstrasse 190, CH-8057 Zürich, Switzerland(Received 15 April 2014; revised manuscript received 14 May 2014; published 23 June 2014)

Hexagonal boron nitride (h-BN) is a prominent member in the growing family of two-dimensional materialswith potential applications ranging from being an atomically smooth support for other two-dimensional materialsto templating growth of molecular layers. We have studied the structure of monolayer h-BN grown by chemicalvapor deposition on Ir(111) by low-temperature scanning tunneling microscopy (STM) and spectroscopy (STS)experiments and state-of-the-art density functional theory (DFT) calculations. The lattice mismatch between theh-BN and Ir(111) surface results in the formation of a moiré superstructure with a periodicity of ∼29 Å and acorrugation of ∼0.4 Å. By measuring the field emission resonances above the h-BN layer, we find a modulationof the work function within the moiré unit cell of ∼0.5 eV. DFT simulations for a 13-on-12 h-BN/Ir(111) unit cellconfirm our experimental findings and allow us to relate the change in the work function to the subtle changesin the interaction between boron and nitrogen atoms and the underlying substrate atoms within the moiré unitcell. Hexagonal boron nitride on Ir(111) combines weak topographic corrugation with a strong work functionmodulation over the moiré unit cell. This makes h-BN/Ir(111) a potential substrate for electronically modulatedthin film and heterosandwich structures.

DOI: 10.1103/PhysRevB.89.235429 PACS number(s): 68.37.Ef, 73.20.−r, 73.22.−f, 81.15.Gh

I. INTRODUCTION

Hexagonal boron nitride (h-BN) is a prominent memberin the growing family of two-dimensional (2D) materials.Isostructural to graphene, while being a wide band gapinsulator, h-BN has found a host of current and potentialapplications. These range from serving as an atomicallysmooth support for other 2D materials to band structureengineering in graphene/h-BN heterostructures to epitaxialgrowth in two-dimensional space [1–3]. Another area ofinterest in h-BN are the so-called boron nitride nanomeshes—epitaxial monolayers of h-BN grown on transition metalsurfaces [4,5]. Various studies have been motivated by theirability to act as a template for bottom-up fabrication tech-niques, while simultaneously providing electronic decouplingfrom the metallic substrate [4,6–13]. The term “nanomesh”highlights the peculiar structure of the h-BN monolayers:Due to the lattice mismatch with the underlying substrate,the atomic registry between the boron and nitrogen atoms andthe metallic surface is periodically modulated, resulting in amoiré pattern formed by regions of stronger h-BN/metal andweaker h-BN/metal interaction.

Topography, work function, and chemical reactivity areperiodically modulated over the moiré pattern, which isthe origin of the templating effect of the h-BN nanomesh.Recent work on extended, self-assembled molecular layers onh-BN/Ir(111) [12] and h-BN/Cu(111) [13] has demonstratedthat these templating capabilities are not only limited tostructural properties but can be extended to the electronicproperties of the overlayer: The work function modulationalong the h-BN moiré unit cell causes an energy shift of themolecular resonances [13], potentially resulting in the local

*peter.liljeroth@aalto.fi

charging of the molecular layer [12]. Thus, monolayers ofh-BN offer a route to grow organic thin films or layered het-erostructures with periodically modulated electronic proper-ties. Three criteria control the properties of the resulting films:(i) structural corrugation of the h-BN layer, (ii) work functionmodulation along the moiré unit cell, and (iii) coherencelength of the moiré pattern. A small structural corrugationfacilitates the growth of defect-free overlayers and sandwichstructures, while a large work function modulation allowsthe patterning of electronic properties. Finally, growing high-quality, large-scale layers requires the h-BN to maintain auniform moiré periodicity and orientation over the entiresample size with a low density of domain boundaries.

A low structural corrugation is generally found inweakly interacting h-BN/metal systems such as h-BN/Cu(111)[11,14,15] or h-BN/Pt(111) [16]. However, the weak in-teraction often leads to the formation of several rotationaldomains [11,17–19], as there is no preferred growth direction.For example on copper, this leads to a variation of themoiré periodicity from �5 up to 14 nm [11], dependingon the rotational alignment. This problem can be bypassedby growing h-BN on strongly interacting transition metals,which typically result in single-domain growth, as well as alarger work function modulation. The prototypical, stronglyinteracting nanomesh systems h-BN/Rh(111) [4,8,20] andh-BN/Ru(0001) [7,20] offer a work function modulation ofup to 0.5 eV along the moiré, however, accompanied bya large structural corrugation of around 1 to 1.5 Å. Ah-BN/metal system combining the advantages of strongly andweakly interacting systems would be of great value for furtherexploration of electronically modulated thin films.

Recent low-energy electron diffraction experiments [21,22]indicate the existence of a preferred orientation of h-BNon Ir(111), while previous density functional theory (DFT)calculations [23] as well as x-ray adsorption and photoelectron

1098-0121/2014/89(23)/235429(8) 235429-1 ©2014 American Physical Society

http://dx.doi.org/10.1103/PhysRevB.89.235429

FABIAN SCHULZ et al. PHYSICAL REVIEW B 89, 235429 (2014)

spectroscopy experiments [16] suggest a weaker interactionof h-BN with Ir(111) than with Rh(111) or Ru(0001). Thus,h-BN/Ir(111) constitutes a promising candidate to combinethe desired properties of preferred orientation of the moirésuperstructure, strong work function modulation, and smallstructural corrugation. Here we show that h-BN grown onIr(111) indeed fulfills these criteria. Low-temperature scan-ning tunneling microscopy (STM) experiments demonstratethat h-BN/Ir(111) can be grown to form large domainsextending across step edges, and with a preferred orientationof the h-BN layer with respect to the substrate lattice, resultingin a low spread of the moiré periodicity. Scanning tunnelingspectroscopy (STS) indicates a modulation of the workfunction along the moiré unit cell of ∼0.5 eV. Our experimentalfindings are complemented by extensive, state-of-the-art DFTcalculations. The computational results confirm the workfunction modulation and indicate a weak structural corrugationof the h-BN layer of 35 pm. Combining the experimentalresults with the simulation of the moiré unit cell, we canexplain the work function modulation as a result of subtlechanges in the registry and interaction between the h-BN andsubstrate atoms.

II. METHODS

Sample preparation. All the experiments were carriedout in an ultrahigh vacuum system with a base pressure of∼10−10 mbar. The (111)-terminated iridium single crystal wascleaned by repeated cycles of sputtering with 1.5 kV neon ions,annealing to 900 ◦C in 5 × 10−7 mbar oxygen and subsequentflashing to 1400 ◦C. Full monolayers of h-BN were grown bythermal cracking of borazine (B3N3H6, Chemos GmbH) at theIr(111) [21] substrate held at a temperature of 1080 ◦C and witha borazine pressure of 2 × 10−8 mbar. With these parameters,the h-BN layer grows with a low nucleation density and formslarge domains with sizes larger than the terrace width ofthe Ir(111) substrate [24]. The resulting h-BN domains arealigned with the substrate lattice, resulting in a very uniformmoiré periodicity of (29.3 ± 0.6) Å. When growing the h-BNat higher substrate temperatures between 1100 and 1200 ◦C,we also find misaligned domains, yielding moire periodicitiesdown to 22 Å. Temperature programed growth of the h-BNlayer by preadsorption of borazine on the sample held at roomtemperatures and subsequent annealing results in a mosaicgrowth [24] of the h-BN, yielding a large spread of rotationaldomains, as well as defects and grain boundaries. Using sampletemperatures significantly above 1200 ◦C suppresses the h-BNgrowth, potentially due to an increase in desorption of nitrogenfrom the Ir surface and solubility of boron into the Ir bulk.

STM measurements. After the preparation, the sample wasinserted into the low-temperature STM (Createc LT-STM) andall subsequent measurements were performed at 5 K. Differen-tial conductance (dI/dV ) spectra were recorded by standardlock-in detection on the tunneling current, while sweeping theapplied sample bias with a peak-to-peak modulation of 20 mVat a frequency of 517 Hz, with the current feedback loopopened at 1 V and 0.25 nA. I (z) spectra were taken at a biasof 50 mV. Field emission resonances (FERs) were measuredwith closed feedback loop at a current setpoint of 0.5 nA.

STM images and dI/dV line spectra were processed usingthe WSxM [25] and SpectraFox [26] softwares, respectively.

DFT calculations. The DFT calculations were performedusing the QuickStep module [27] of the CP2K package(http://www.CP2K.org/), where the Kohn-Sham orbitals areexpanded in the basis of Gaussian functions (here DZVP-MOLOPT-SR-GTH for B and N and DZVP-MOLOPT-SR-GTH-q17 for Ir [28]), and plane waves up to a cut-off energy of700 Ry, REL_CUTOFF of 70 Ry, for the density. Generalizedgradient approximation (GGA) revPBE [29] was employed asthe exchange-correlation functional, and the missing Londondispersion incorporated using the semiempirical DFT-D3formalism [30]. The surface was modeled using the slabapproach with four layers of the substrate of which the twoat the bottom were held fixed during the geometry relaxation,and the h-BN layer adsorbed only on one side of the slab. Thelength of the cell was 40 Å in order to decouple the two sidesof the slab from each other. Only the � point was used in theevaluation of the integrals over the first Brillouin zone, andthe bulk lattice constant of 3.801 Å obtained with revPBE-D3was used. The occupation numbers were broadened using theFermi-Dirac distribution at 300 K around the Fermi energy.The STM images were simulated using the Tersoff-Hamannmodel [31] with an s wave tip. Further details on the methodcan be found in Ref. [20]

III. RESULTS AND DISCUSSION

Figure 1(a) shows a large-scale STM image of monolayer h-BN grown on the Ir(111) surface by chemical vapor deposition(CVD) using borazine as the precursor (see Sec. II fordetails). A single domain extends over several monatomicsteps, indicating the high quality of the h-BN layer as itextends over step edges in a carpetlike fashion [24,32,33].The moiré superstructure due to the lattice mismatch (Ir(111):a = 2.714 Å [34] and h-BN: a = 2.505 Å [35]) is highlightedin Fig. 1(b); it is formed by depressions arranged in a hexagonallattice. The periodicity of the superstructure is indicated inthe right panel of Fig. 1(b) by a hexagonal grid overlayedonto the STM image. Throughout this article we will use thecommon terminology and refer to the depressions of the moiréas pores and to the surrounding regions as wires. However, itis important to note that the h-BN forms a continuous, closedmonolayer without any voids. The moiré unit cell as wellas the atomic unit cell of the h-BN are depicted in the highresolution STM topograph in Fig. 1(c). As can be seen, the twounit cells are aligned, without any appreciable rotation withrespect to each other. Since the angle between the moiré andh-BN lattice vectors represents a roughly tenfold magnificationof the rotation between the lattice of h-BN and Ir(111) [36],we conclude that the misalignment between the two atomiclattices is less than ±0.5◦. When growing the h-BN layer at asubstrate temperature of 1080 ◦C, we only find aligned h-BNwith a moiré periodicity of (29.3 ± 0.6) Å. The small deviationfrom the theoretical moiré periodicity of 32.5 Å suggests thatthe h-BN lattice is strained by approximately 0.8%, similar tothe case of graphene on Ir(111) [37,38].

A notable feature in the STM image is a bright rim aroundthe pores. This rim appears only at a certain sample biasrange and is a direct consequence of the different atomic

235429-2

http://www.CP2K.org/

EPITAXIAL HEXAGONAL BORON NITRIDE ON Ir(111): . . . PHYSICAL REVIEW B 89, 235429 (2014)

3rd Ir

2nd Ir

1st Ir

N

B

3.00

3.05

3.10

3.15

3.20

3.25

Hei

ght a

bove

1st

Ir la

yer (

Å)

(d)

NfccBtop

NhcpBfcc

(b)

40 nm

2 nm

(c)

5 nm

(a)

PW

NtopBhcp

FIG. 1. (Color online) Moiré superstructure of hexagonal boronnitride on Ir(111). (a) STM overview image of the sample showinga single h-BN domain extending over four monatomic steps. Theblack and white color scale is adjusted to repeatedly cover theheight of two terraces. The inset is a fast Fourier transform of theimage, showing six sharp spots indicative of a single rotational moirédomain. (b) Zoom-in on the moiré pattern (left), highlighting itshexagonal periodicity with an overlayed grid (right). (c) Atomicallyresolved STM image; the moiré and atomic unit cells are indicatedin black and white, respectively. Pore and wire regions are markedby a blue “P” and a red “W”, respectively. (d) DFT simulation ofh-BN on Ir(111). The moiré unit cell as well as regions where Band N atoms occupy high-symmetry positions with respect to theIr lattice are indicated. Feedback parameters: (a) 1.66 V, 0.31 nA;(b) −1.71 V, 0.40 nA; and (c) 0.20 V, 3.00 nA.

registry within the moiré unit cell, leading to a modulationof the electronic properties of the h-BN overlayer, as will beshown later. On the basis of such high-resolution images andprevious results by Orlando et al. [21], we have performed adispersion-corrected DFT simulation for a 13×13 on 12×12h-BN/Ir(111) unit cell, i.e., along the vector of the moiré unitcell 13 h-BN unit cells occupy 12 substrate unit cells (seeSec. II for details). The fully relaxed theoretical h-BN/Ir(111)structure shown in Fig. 1(d) reproduces the pore and wirepattern observed in our STM experiments. From the simulation

we find the depressions of the moiré corresponding to a registrywith the center of the B-N hexagon over a fcc hollow site ofthe Ir(111) lattice, the nitrogen atom sitting on a top site,and the boron atom on a hcp hollow site (BhcpNtop). Theminimum distance between the topmost iridium layer and theh-BN lattice in this configuration is 2.95 Å. The maximumdistance between the iridium and the h-BN is found on thewire, when N atoms occupy hcp hollow sites and B atomsfcc hollow sites (BfccNhcp), the center of the hexagon thusbeing on top the underlying Ir atoms. At this registry, theh-BN-Ir distance is 3.30 Å, giving a total corrugation withinthe moiré superstructure of 35 pm. When the nitrogen sits onfcc hollow sites and the boron on top sites (BtopNfcc, the centerof the hexagon thus on a hcp hollow site), the distance to theiridium layer is slightly smaller, i.e., 3.20 Å. On the rim of thepore at the transition towards the wire, boron atoms occupypredominantly bridge positions and nitrogens off-center toppositions (BbriNtop). While the correspondence between theatomic registry and the different areas of the moiré unit cellis similar to previous findings for other h-BN/metal systems[20], the corrugation is much smaller than in the stronglyinteracting systems such as Rh(111) [4,8] or Ru(0001) [7],where the difference between pore and wire regions is around1 Å [20]. With 35 pm, the corrugation of the h-BN/Ir(111)system is comparable to the one found for graphene on Ir(111)(∼45 pm) [39]. In addition, the calculated minimum heightof the h-BN monolayer above the Ir surface is close to thevalue of the interlayer distance in bulk h-BN of 3.33 Å [35],while for h-BN/Rh(111) and h-BN/Ru(0001) the minimumheight is between 2 and 2.5 Å [20]. This is clear indicationthat the interaction between h-BN and the Ir(111) surface ismuch weaker compared to the prototypical nanomesh systems.However, these subtle differences in atomic registry alongthe moiré unit cell give rise to noticeable variations of theelectronic properties along the h-BN layer.

Figure 2(a) compares differential conductance (dI/dV )spectra measured on the bare Ir(111) and on the pore andwire regions of the h-BN moiré. The reference spectrum ofthe bare iridium shows a steplike increase in the conductivityat ∼−380 mV due to the holelike surface state present at the(111)-terminated Ir surface [40,41]. Metallic surface states arewell known to be very sensitive to any kind of adsorbates.Depending on the nature and the strength of the interactionbetween surface and adsorbate, the binding energy of thesurface state can shift [42–46], its onset can broaden [44],and its intensity can be attenuated [43,46–48] or eventually becompletely quenched [45,49–51]. The dI/dV spectrum takenon the wire region of the h-BN moiré exhibits a steplike featurearound ∼−360 mV, which we assign to the surface state. Thefact that the surface state survives underneath the wire regionand does not shift significantly confirms the weak interactionof this part of the moiré unit cell with the Ir(111) surface.We note, however, that the onset is broadened, suggesting adecreased lifetime of holes in the surface state [52]. In contrast,the spectrum taken on the pore does not show any featuresrelated to the surface state. It thus appears to be quenched (orat least strongly shifted) by the stronger interaction of the h-BNlayer with the metallic surface on the pore region of the moiréunit cell. Instead, we find a sharp rise in the dI/dV signal atpositive bias, starting around 1 V. The evolution in differential

235429-3

FABIAN SCHULZ et al. PHYSICAL REVIEW B 89, 235429 (2014)

dI/dVhigh

low

Sample bias (V)

Late

ral p

ositi

on

-2 -1 0 21-2 -1 0 10

1

2

dI/d

V (

a.u.

)

Sample bias (V)

0 30 60 90 120

0

3

6

9

12

1.7 V

~1.0 Å

2.1 V

App

aren

t hei

ght (

Å)

Distance (Å)

~ -0.5 Å

~2.8 Å

~2.8 Å

0.9 V

1.3 V

-0.1 V

-0.5 V

0.1 V

0.5 V

-1.7 V

-2.1 V

-1.3 V

-0.9 V

Ir(111)

W

P

2 nm

2.1 V

-2.1 V

-0.1 V

(a) (b)

(c) (d)

2 nm

FIG. 2. (Color online) Experimental electronic structure of h-BNon Ir(111). (a) dI/dV spectra taken on the pore and on the wire ofthe moiré. The inset is an STM image indicating the positions for thepoint spectra as well as for the line spectra in (b). (b) Color-codeddI/dV spectra taken along a line connecting two pores. (c) and (d)Bias-dependent STM contrast of the h-BN layer. Line profiles in(c) correspond to green lines in (d). Feedback parameters: Inset (a)2.10 V, 0.25 nA; (d) 0.25 nA for all images, sample bias as indicated.

conductance along the moiré unit cell is depicted in Fig. 2(b),where a series of individual dI/dV spectra is shown as afunction of lateral position in a two-dimensional color plot.The strong increase at positive bias is observed exclusively atthe pore, thus its origin must lie in the interaction of the h-BNwith the metal substrate.

The different features of the dI/dV spectra are alsoreflected in the STM image contrast. Figure 2(c) shows STMline profiles of the h-BN moiré extracted from images takenat different sample biases showing clean Ir(111) as well as anh-BN island. The series goes from −2.1 to 2.1 V and the profilecrosses two moiré unit cells, as indicated by the green linesin the STM images in Fig. 2(d). At negative bias, the poresappear as depressions with an apparent depth of 0.3–0.5 Å, inagreement with the actual topographic corrugation obtainedfrom the DFT simulation [Fig. 1(d)]. However, at largepositive bias the moiré contrast inverts and the pores appear asprotrusions with an apparent height of up to ∼1 Å with respectto the wire region. This can be directly related to the increasedDOS on the pore region at high bias as depicted in the spectrain Figs. 2(a) and 2(b). At low biases, a bright rim appearsaround the pores with an apparent height larger than the wire

(a)

(b)-3.0 V -0.5 V 3.0 V

-10 -8 -6 -4 -2 0 2 4 6 8 10-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3

without Ir(111)

PDO

S in

pz p

er a

tom

(a.u

.)

Energy (eV)

Nhcp B fcc Nfcc B top Ntop B hcp

Ntop B bri

with Ir(111)

FIG. 3. (Color online) Theoretical electronic structure of theh-BN monolayer on Ir(111). (a) The projected density of states ofthe pz orbitals of boron and nitrogen as obtained from the DFTcalculation. (b) Simulated STM images at different bias voltages.

or the pore. Note that the apparent height of the wire regionwith respect to the clean Ir(111) does not show any significantchange over the entire bias region, being around ∼2.8 Å. Thisis in rough agreement with the h-BN-Ir(111) distance on thewire given by DFT of 3.3 Å [Fig. 1(d)].

To further elucidate the origin of this contrast reversal, wehave plotted in Fig. 3(a) the projected density of the states(PDOS) of the pz orbital for boron and nitrogen atoms in theh-BN layer with (“positive” PDOS) and without (“negative”PDOS) the Ir(111) surface as obtained from our DFT cal-culations. The PDOS is split into different atomic registrieswith respect to the iridium surface, corresponding to differentareas of the moiré unit cell. As expected, irrespective of thepresence of the surface, the occupied states are dominated bythe nitrogen atoms (full lines) and the unoccupied ones by theboron atoms (dashed lines). At the pore and at its rim, wherethe registries are BhcpNtop and BbriNtop, the entire PDOS isshifted to lower energies, i.e., the onset of the conduction bandis observed at lower energies than at the wire region, causingthe pores to appear bright at positive bias. As the valenceband is shifted downwards as well, it becomes accessible onlyat larger negative bias compared to the wire, resulting in thepores being imaged as depressions and thus reversing the moiréSTM contrast. Apart from the downward shift of the bands onthe pore region, there is finite PDOS around the Fermi energyfor the h-BN in Ir(111), indicating a partial—albeit small—hybridization of h-BN pz states with electronic states of theunderlying metal substrate (most likely with the partially filledd states of the iridium [16,21,23]) in agreement with previousexperimental findings [16]. Again, this effect is the strongestfor the registries corresponding to the pore and rim regionsof the moiré, in accordance with the increased interactionwith the substrate. At small energies around zero, the rimand the pore show the largest DOS, which in conjunction withthe actual topographic corrugation causes the rim to appearthe brightest at small biases. We can reproduce the observed

235429-4

EPITAXIAL HEXAGONAL BORON NITRIDE ON Ir(111): . . . PHYSICAL REVIEW B 89, 235429 (2014)

contrast changes by using our DFT results to simulate STMimages, following the Tersoff-Hamann model [31]. Figure 3(b)depicts a set of such simulations, for sample biases of −3.0,−0.5, and 3.0 V. The images at large negative and largepositive bias yield excellent agreement with the experimentalobservations, as the appearance of the pore switches from adepression to a protrusion. Also the bright rim around the poreat low voltages is well reproduced by our STM simulation.

These variations in the electronic structure of the h-BN layersuggest a possible modification of the local tunneling barrieralong the moiré unit cell. The tunneling barrier is directlyrelated to the local work function or surface potential, whichcan be probed with high spatial resolution by measuring fieldemission resonances in the STM junction [53,54]. FERs, alsoknown as Gundlach oscillations [55], arise in the regime ofFowler-Nordheim tunneling [56], i.e., when the applied biasis larger than the sample work function and thus, the tip Fermilevel is above the vacuum level of the sample. The trapezoidalpotential due to the drop of the bias voltage along the tunnelingjunction can give rise to hydrogenlike electronic resonancesconfined in the vacuum junction by the sample surface and theclassical turning point at that trapezoidal potential, as depictedschematically in Fig. 4(a). Qualitatively, these resonances canbe understood as image potential states under an externalelectric field. As the energy of these resonances depends onthe local work function of the sample [57], FERs have foundwide application in the STM community to map work functionchanges, in particular of thin films grown on metal substratessuch as oxide films [58,59], thin insulating layers of NaCl[60,61], and CuN [62] or monolayers of graphene [63] andh-BN [11]. To a first approximation, the sample work functionis given by the energy of the first FER. Figure 4(b) shows FERspectra measured at different parts of the h-BN moiré as wellas on the clean Ir(111) for comparison.

Tip Sample

EnergyPotential

Evac

EF,T

EF,S

e×VS

T

(a)(b)

(c)

2 nm

(d)

2 4 6 8 100

1

2

3

4

dI/d

V (

a.u.

)

Sample bias (V)

-1 0 1 20.001

0.01

0.1

1

Nor

mal

ized

cur

rent

Relative tip-sample approach (Å)

Ir(111)

Ir(111)

FIG. 4. (Color online) Field emission resonances and I (z) spec-troscopy on h-BN/Ir(111). (a) STM junction under high bias in theFER regime. (b) FER and (c) I (z) point spectroscopy measured on theclean Ir(111) (black) and on different parts of the h-BN moiré. FERspectra vertically offset for clarity. (d) STM topography indicatingthe location of the point spectra in (b) and (c). Feedback parameters:(d) 0.20 V, 1.00 nA.

On the bare iridium, the first FER appears at ∼5.8 V, ingood agreement with the work function of Ir(111) of 5.76 eV[64]. In contrast, the h-BN yields its first FER at ∼4.2 to4.4 V, indicating a strong reduction of the work functionon the h-BN overlayer of around 1.4 eV. This overall workfunction reduction is comparable with previously reportedvalues for other h-BN/metal systems [7,11,18]. Interestingly,the first three FERs of the h-BN layer show some internalstructure, being actually composed of three subpeaks, whoserelative intensities vary depending on the area within the moiréunit cell. Such an effect has been observed previously whenmeasuring FERs over the moiré of one monolayer of FeO onPt(111) and was attributed to contributions from different partsof the moiré unit cell [58]. It has been pointed out that sinceFERs are measured in a closed feedback configuration and atlarge bias (thus at a large tip-sample distance) the effectivearea that is probed can be in the order of 100 Å2 [54,58]. Thesubpeaks are most clearly resolved in the second FER of the h-BN. In fact, it has been shown that the energetic position of thesecond FER is a good measure to determine relative shifts ofthe surface potential on the local scale, as it is less influenced bythe image potential at the sample surface [65]. Inspecting theinternal structure of the second FER for the different parts ofthe h-BN moiré, we find the three peaks being located at ∼6.0,6.3, and 6.6 V. The subpeak of lowest energy has its maximumintensity at the pore, while the peak of highest energy showsmaximum intensity on the wire region, indicating a reductionof the work function when going from the wire to the pore. Thisresult is supported by measurements of the local barrier heightsusing I (z) spectroscopy. The tunneling current decreases ex-ponentially with increasing tip-sample distance, whereby thedecay constant is proportional to the square root of the apparentbarrier height. Normalized I (z) spectra, taken on clean Ir(111)and on different parts of the h-BN moiré, are plotted inFig. 4(c); it clearly can be seen that on both the pore and wire,the current decays much more slowly than on the clean iridium,indicating a lower apparent tunneling barrier. Approximatingthe barrier height �b as the average work function of the tip-sample system, i.e., �b = (�t + �s)/2 [66], we can deducefrom an exponential fit and via ��s = 2(�Ir − �hBN) [66]an overall work function reduction with respect to the iridiumsurface of ∼1.6 eV for the wire and ∼1.2 eV for the pore (notethat �Ir and �hBN refer to the potential barriers as determinedfrom the exponential fit). Combining the results of the FER andI (z) measurements, we find a modulation of the work functionwithin the moiré of roughly 0.5 eV.

For a more detailed mapping of the work function varia-tions, we measured FER spectra along the high-symmetry lineof the moiré unit cell connecting two next-nearest neighborpores with each other [Fig. 5(a)]. The result is displayed ascolor-coded dI/dV intensity as a function of lateral positionand applied sample bias in Fig. 5(b). At the pores, the secondFER yields its highest intensity at ∼6.2 V, with little variationalong the area of the pore. On the wire region, the maximumintensity is found between ∼6.6 and 6.7 V, with a slight asym-metry between the upper and lower part of the wire. Interest-ingly, the transition from low work function at the pore to highwork function at the wire appears rather sudden, indicated bythe coexistence region of the two corresponding peaks insteadof a smooth shift of the second FER towards higher energy.

235429-5

FABIAN SCHULZ et al. PHYSICAL REVIEW B 89, 235429 (2014)

2 nm 4.2

4.6

4.4

4.3

4.5

Local work function (eV

)

dI/dVhigh

low

Sample bias (V)

Late

ral p

ositi

on2 4 6 108

(a) (b) (c)

FIG. 5. (Color online) Mapping the work function changes overthe moiré unit cell. (a) STM topography image showing the locationof the FER line spectra plotted in (b). (b) Color-coded dI/dV FERspectra taken along the line indicated in (a). (c) Theoretical workfunction changes over the moiré unit cell computed from DFT.Feedback parameters: (a) 0.20 V, 1.00 nA.

To compare the experimentally measured variations in thelocal surface potential with our DFT calculations, we haveplotted the calculated Hartree potential above the h-BN layerin Fig. 5(c), which is approximately equal to the local workfunction. The simulation confirms that the pores of the h-BNmoiré yield the lowest work function of ∼4.2 eV; the highestwork function of ∼4.6 eV is found at the wire region thatcorresponds to the BfccNhcp registry, while the BtopNfcc wireregion has a slightly lower work function. Thus, the variationof the work function follows the modulation in the interactionstrength between the different regions of the h-BN moiréand the iridium substrate, allowing us to relate the observedchanges to the calculated PDOS of the B and N pz orbitals[Fig. 3(a)]. First, we note that the overall reduction of thework function—confirmed by our DFT calculations to be morethan 1 eV—can be explained by the “push-back” or “pillow”effect [67,68]: Due to the large spill-out of the electronicwave functions at a metal surface, an interface dipole layerpointing into the vacuum is formed [69]. This dipole layer isknown to have a significant contribution to the work functionof metals (as it points towards the vacuum, it increases themagnitude of the work function) [69–71]. However, uponformation of the h-BN layer, the wave function spill-out isstrongly reduced (“push-back”) due to the Pauli exclusionprinciple, resulting in a reduction of the work function. Now,to explain the variations of the work function within the h-BNmoiré unit cell, we recall that the PDOS plotted in Fig. 3(a)indicates a hybridization of pz orbitals of the nitrogen atomswith states of the underlying metal substrate. This implies aredistribution of electron density from the h-BN layer acrossthe interface towards the metal substrate, i.e., the h-BN layerbecomes slightly positively charged. As a result, an interfacedipole pointing towards the metal substrate is formed and

reduces the work function further. This hybridization effectis the strongest at the pore where N occupies top positions andthe interaction between h-BN and Ir(111) is maximum, thusthe work function is the lowest on the pore. On the wire, inparticular on the least interacting regions where the registry isBfccNhcp and BtopNfcc, the hybridization is minimal and thus,this is the region with the largest work function within themoiré unit cell. Overall, the calculated work function yieldsvery good agreement—qualitatively and quantitatively—withour FER measurements.

IV. CONCLUSIONS

In summary, we have provided a detailed description of thestructure of monolayer h-BN grown on Ir(111) at the atomiclevel. Due to the lattice mismatch between the h-BN and theIr(111), a moiré superstructure with a periodicity of ∼29 Å anda corrugation of ∼0.4 Å is formed. The strongly interactingpores of the moiré corresponds to a BhcpNtop registry, whilethe regions on the wire with a BfccNhcp registry have theweakest interaction with the substrate. The pz orbitals of thenitrogen atoms partially hybridize with the metal substrate. Asthe magnitude of the hybridization depends on the interactionstrength and thus on the atomic registry, it gives rise to amodulation of the work function within the moiré unit cell of∼0.5 eV. Overall, the h-BN layer reduces the work functionof the Ir(111) substrate by more than 1 eV. Our results are inline with previous findings that the strength of the chemicalinteraction at the interface of h-BN and Ir(111) should beless than for Rh(111) or Ru(0001) but more than for Pt(111)[16]. Therefore, the moiré pattern formed by h-BN/Ir(111)combines the advantages found in a strongly interactingh-BN/metal system of a large work function modulation andsingle domain growth with a low structural corrugation, thelatter usually characteristic for weakly interacting h-BN/metalsystems. This makes it a superior candidate for the bottom-up fabrication of electronically modulated thin films andheterosandwich structures.

ACKNOWLEDGMENTS

This research made use of the Aalto NanomicroscopyCenter (Aalto NMC) facilities and was supported by theEuropean Research Council (ERC-2011-StG No. 278698“PRECISE-NANO”), the Academy of Finland (Centre ofExcellence in Low Temperature Quantum Phenomena andDevices No. 250280), and the Finnish Academy of Scienceand Letters. Computing time was awarded by CSCS (CentroSvizzero di Calcolo Scientifico) under the Project No. s425.

[1] C. R. Dean, A. F. Young, I. Meric, C. Lee, L. Wang, S. Sorgenfrei,K. Watanabe, T. Taniguchi, P. Kim, K. L. Shepard, and J. Hone,Nat. Nanotechnol. 5, 722 (2010).

[2] B. Hunt, J. D. Sanchez-Yamagishi, A. F. Young, M. Yankowitz,B. J. LeRoy, K. Watanabe, T. Taniguchi, P. Moon, M. Koshino,P. Jarillo-Herrero, and R. C. Ashoori, Science 340, 1427 (2013).

[3] L. Liu, J. Park, D. A. Siegel, K. F. McCarty, K. W. Clark,W. Deng, L. Basile, J. C. Idrobo, A.-P. Li, and G. Gu, Science343, 163 (2014).

[4] M. Corso, W. Auwärter, M. Muntwiler, A. Tamai, T. Greber,and J. Osterwalder, Science 303, 217 (2004).

[5] R. Laskowski, P. Blaha, T. Gallauner, and K. Schwarz, Phys.Rev. Lett. 98, 106802 (2007).

[6] M. Corso, T. Greber, and J. Osterwalder, Surf. Sci. 577, L78(2005).

[7] A. Goriachko, Y. He, M. Knapp, H. Over, M. Corso, T. Brugger,S. Berner, J. Osterwalder, and T. Greber, Langmuir 23, 2928(2007).

235429-6

http://dx.doi.org/10.1038/nnano.2010.172http://dx.doi.org/10.1038/nnano.2010.172http://dx.doi.org/10.1038/nnano.2010.172http://dx.doi.org/10.1038/nnano.2010.172http://dx.doi.org/10.1126/science.1237240http://dx.doi.org/10.1126/science.1237240http://dx.doi.org/10.1126/science.1237240http://dx.doi.org/10.1126/science.1237240http://dx.doi.org/10.1126/science.1246137http://dx.doi.org/10.1126/science.1246137http://dx.doi.org/10.1126/science.1246137http://dx.doi.org/10.1126/science.1246137http://dx.doi.org/10.1126/science.1091979http://dx.doi.org/10.1126/science.1091979http://dx.doi.org/10.1126/science.1091979http://dx.doi.org/10.1126/science.1091979http://dx.doi.org/10.1103/PhysRevLett.98.106802http://dx.doi.org/10.1103/PhysRevLett.98.106802http://dx.doi.org/10.1103/PhysRevLett.98.106802http://dx.doi.org/10.1103/PhysRevLett.98.106802http://dx.doi.org/10.1016/j.susc.2005.01.015http://dx.doi.org/10.1016/j.susc.2005.01.015http://dx.doi.org/10.1016/j.susc.2005.01.015http://dx.doi.org/10.1016/j.susc.2005.01.015http://dx.doi.org/10.1021/la062990thttp://dx.doi.org/10.1021/la062990thttp://dx.doi.org/10.1021/la062990thttp://dx.doi.org/10.1021/la062990t

EPITAXIAL HEXAGONAL BORON NITRIDE ON Ir(111): . . . PHYSICAL REVIEW B 89, 235429 (2014)

[8] S. Berner, M. Corso, R. Widmer, O. Groening, R. Laskowski,P. Blaha, K. Schwarz, A. Goriachko, H. Over, S. Gsell, M.Schreck, H. Sachdev, T. Greber, and J. Osterwalder, Angew.Chem. Int. Ed. 46, 5115 (2007).

[9] H. Dil, J. Lobo-Checa, R. Laskowski, P. Blaha, S. Berner,J. Osterwalder, and T. Greber, Science 319, 1824 (2008).

[10] S. Kahle, Z. Deng, N. Malinowski, C. Tonnoir, A. Forment-Aliaga, N. Thontasen, G. Rinke, D. Le, V. Turkowski, T. S.Rahman, S. Rauschenbach, M. Ternes, and K. Kern, Nano Lett.12, 518 (2012).

[11] S. Joshi, D. Ecija, R. Koitz, M. Iannuzzi, A. P. Seitsonen,J. Hutter, H. Sachdev, S. Vijayaraghavan, F. Bischoff,K. Seufert, J. V. Barth, and W. Auwaerter, Nano Lett. 12, 5821(2012).

[12] F. Schulz, R. Drost, S. K. Hämäläinen, and P. Liljeroth, ACSNano 7, 11121 (2013).

[13] S. Joshi, F. Bischoff, R. Koitz, D. Ecija, K. Seufert, A. P.Seitsonen, J. Hutter, K. Diller, J. I. Urgel, H. Sachdev, J. V.Barth, and W. Auwärter, ACS Nano 8, 430 (2014).

[14] A. Preobrajenski, A. Vinogradov, and N. Mårtensson, Surf. Sci.582, 21 (2005).

[15] S. Roth, F. Matsui, T. Greber, and J. Osterwalder, Nano Lett. 13,2668 (2013).

[16] A. B. Preobrajenski, M. A. Nesterov, M. L. Ng, A. S.Vinogradov, and N. Martensson, Chem. Phys. Lett. 446, 119(2007).

[17] F. Müller, K. Stöwe, and H. Sachdev, Chem. Mater. 17, 3464(2005).

[18] M. Morscher, M. Corso, T. Greber, and J. Osterwalder, Surf.Sci. 600, 3280 (2006).

[19] F. Müller, S. Hüfner, H. Sachdev, R. Laskowski, P. Blaha, andK. Schwarz, Phys. Rev. B 82, 113406 (2010).

[20] J. G. Diaz, Y. Ding, R. Koitz, A. P. Seitsonen, M. Iannuzzi, andJ. Hutter, Theor. Chem. Acc. 132, 1350 (2013).

[21] F. Orlando, R. Larciprete, P. Lacovig, I. Boscarato, A. Baraldi,and S. Lizzit, J. Phys. Chem. C 116, 157 (2012).

[22] D. Usachov, A. Fedorov, O. Vilkov, V. K. Adamchuk, L. V.Yashina, L. Bondarenko, A. A. Saranin, A. Grüneis, and D. V.Vyalikh, Phys. Rev. B 86, 155151 (2012).

[23] R. Laskowski, P. Blaha, and K. Schwarz, Phys. Rev. B 78,045409 (2008).

[24] J. Lu, P. S. E. Yeo, Y. Zheng, H. Xu, C. K. Gan, M. B. Sullivan,A. Castro Neto, and K. P. Loh, J. Am. Chem. Soc. 135, 2368(2013).

[25] I. Horcas, R. Fernandez, J. M. Gomez-Rodriguez, J. Colchero,J. Gomez-Herrero, and A. M. Baro, Rev. Sci. Instrum. 78,013705 (2007).

[26] M. Ruby, SpectraFox, http://www.spectrafox.com.[27] J. VandeVondele, M. Krack, F. Mohamed, M. Parrinello,

T. Chassaing, and J. Hutter, Comput. Phys. Commun. 167, 103(2005).

[28] J. VandeVondele and J. Hutter, J. Chem. Phys. 127, 114105(2007).

[29] Y. K. Zhang and W. T. Yang, Phys. Rev. Lett. 80, 890(1998).

[30] S. Grimme, J. Antony, S. Ehrlich, and H. Krieg, J. Chem. Phys.132, 154104 (2010).

[31] J. Tersoff and D. R. Hamann, Phys. Rev. B 31, 805 (1985).[32] J. Coraux, A. T. N‘Diaye, C. Busse, and T. Michely, Nano Lett.

8, 565 (2008).

[33] P. W. Sutter, J.-I. Flege, and E. A. Sutter, Nat. Mater. 7, 406(2008).

[34] H. P. Singh, Acta Crystallogr. Sect. A 24, 469 (1968).[35] R. S. Pease, Acta Crystallogr. 5, 356 (1952).[36] A. T. N’Diaye, J. Coraux, T. N. Plasa, C. Busse, and T. Michely,

New J. Phys. 10, 043033 (2008).[37] N. Blanc, J. Coraux, C. Vo-Van, A. T. N’Diaye, O. Geaymond,

and G. Renaud, Phys. Rev. B 86, 235439 (2012).[38] H. Hattab, A. T. N’Diaye, D. Wall, C. Klein, G. Jnawali,

J. Coraux, C. Busse, R. van Gastel, B. Poelsema, T. Michely,F.-J. Meyer zu Heringdorf, and M. Horn-von Hoegen, Nano Lett.12, 678 (2012).

[39] S. K. Hämäläinen, M. P. Boneschanscher, P. H. Jacobse, I. Swart,K. Pussi, W. Moritz, J. Lahtinen, P. Liljeroth, and J. Sainio, Phys.Rev. B 88, 201406 (2013).

[40] A. Varykhalov, D. Marchenko, M. R. Scholz, E. D. L. Rienks,T. K. Kim, G. Bihlmayer, J. Sánchez-Barriga, and O. Rader,Phys. Rev. Lett. 108, 066804 (2012).

[41] S. J. Altenburg, J. Kröger, T. O. Wehling, B. Sachs, A. I.Lichtenstein, and R. Berndt, Phys. Rev. Lett. 108, 206805(2012).

[42] S. G. Louie, Phys. Rev. Lett. 40, 1525 (1978).[43] P. Soukiassian, R. Riwan, J. Lecante, E. Wimmer, S. R. Chubb,

and A. J. Freeman, Phys. Rev. B 31, 4911 (1985).[44] J. Y. Park, U. D. Ham, S. J. Kahng, Y. Kuk, K. Miyake, K. Hata,

and H. Shigekawa, Phys. Rev. B 62, R16341(R) (2000).[45] H. Hövel, B. Grimm, and B. Reihl, Surf. Sci. 477, 43 (2001).[46] F. Forster, G. Nicolay, F. Reinert, D. Ehm, S. Schmidt, and

S. Hüfner, Surf. Sci. 532, 160 (2003).[47] B. T. Jonker, J. F. Morar, and R. L. Park, Phys. Rev. B 24, 2951

(1981).[48] N. Nicoara, E. Roman, J. M. Gomez-Rodriguez, J. A. Martin-

Gago, and J. Mendez, Organic Electron. 7, 287 (2006).[49] W. Eberhardt, S. G. Louie, and E. W. Plummer, Phys. Rev. B

28, 465 (1983).[50] C. T. Tzeng, W. S. Lo, J. Y. Yuh, R. Y. Chu, and K. D. Tsuei,

Phys. Rev. B 61, 2263 (2000).[51] N. Gonzalez-Lakunza, I. Fernandez-Torrente, K. J. Franke,

N. Lorente, A. Arnau, and J. I. Pascual, Phys. Rev. Lett. 100,156805 (2008).

[52] J. Kliewer, R. Berndt, E. V. Chulkov, V. M. Silkin, P. M.Echenique, and S. Crampin, Science 288, 1399 (2000).

[53] G. Binnig, K. H. Frank, H. Fuchs, N. Garcia, B. Reihl, H. Rohrer,F. Salvan, and A. R. Williams, Phys. Rev. Lett. 55, 991 (1985).

[54] R. S. Becker, J. A. Golovchenko, and B. S. Swartzentruber,Phys. Rev. Lett. 55, 987 (1985).

[55] K. H. Gundlach, Solid State Electron. 9, 949 (1966).[56] R. H. Fowler and L. Nordheim, Proc. R. Soc. A 119, 173 (1928).[57] O. Y. Kolesnychenko, Y. A. Kolesnichenko, O. Shklyarevskii,

and H. van Kempen, Physica B 291, 246 (2000).[58] E. D. L. Rienks, N. Nilius, H. P. Rust, and H. J. Freund, Phys.

Rev. B 71, 241404 (2005).[59] T. König, G. H. Simon, H. P. Rust, and M. Heyde, J. Phys. Chem.

C 113, 11301 (2009).[60] M. Pivetta, F. Patthey, M. Stengel, A. Baldereschi, and W. D.

Schneider, Phys. Rev. B 72, 115404 (2005).[61] H.-C. Ploigt, C. Brun, M. Pivetta, F. Patthey, and W.-D.

Schneider, Phys. Rev. B 76, 195404 (2007).[62] C. D. Ruggiero, T. Choi, and J. A. Gupta, Appl. Phys. Lett. 91,

253106 (2007).

235429-7

http://dx.doi.org/10.1002/anie.200700234http://dx.doi.org/10.1002/anie.200700234http://dx.doi.org/10.1002/anie.200700234http://dx.doi.org/10.1002/anie.200700234http://dx.doi.org/10.1126/science.1154179http://dx.doi.org/10.1126/science.1154179http://dx.doi.org/10.1126/science.1154179http://dx.doi.org/10.1126/science.1154179http://dx.doi.org/10.1021/nl204141zhttp://dx.doi.org/10.1021/nl204141zhttp://dx.doi.org/10.1021/nl204141zhttp://dx.doi.org/10.1021/nl204141zhttp://dx.doi.org/10.1021/nl303170mhttp://dx.doi.org/10.1021/nl303170mhttp://dx.doi.org/10.1021/nl303170mhttp://dx.doi.org/10.1021/nl303170mhttp://dx.doi.org/10.1021/nn404840hhttp://dx.doi.org/10.1021/nn404840hhttp://dx.doi.org/10.1021/nn404840hhttp://dx.doi.org/10.1021/nn404840hhttp://dx.doi.org/10.1021/nn406024mhttp://dx.doi.org/10.1021/nn406024mhttp://dx.doi.org/10.1021/nn406024mhttp://dx.doi.org/10.1021/nn406024mhttp://dx.doi.org/10.1016/j.susc.2005.02.047http://dx.doi.org/10.1016/j.susc.2005.02.047http://dx.doi.org/10.1016/j.susc.2005.02.047http://dx.doi.org/10.1016/j.susc.2005.02.047http://dx.doi.org/10.1021/nl400815whttp://dx.doi.org/10.1021/nl400815whttp://dx.doi.org/10.1021/nl400815whttp://dx.doi.org/10.1021/nl400815whttp://dx.doi.org/10.1016/j.cplett.2007.08.028http://dx.doi.org/10.1016/j.cplett.2007.08.028http://dx.doi.org/10.1016/j.cplett.2007.08.028http://dx.doi.org/10.1016/j.cplett.2007.08.028http://dx.doi.org/10.1021/cm048629ehttp://dx.doi.org/10.1021/cm048629ehttp://dx.doi.org/10.1021/cm048629ehttp://dx.doi.org/10.1021/cm048629ehttp://dx.doi.org/10.1016/j.susc.2006.06.016http://dx.doi.org/10.1016/j.susc.2006.06.016http://dx.doi.org/10.1016/j.susc.2006.06.016http://dx.doi.org/10.1016/j.susc.2006.06.016http://dx.doi.org/10.1103/PhysRevB.82.113406http://dx.doi.org/10.1103/PhysRevB.82.113406http://dx.doi.org/10.1103/PhysRevB.82.113406http://dx.doi.org/10.1103/PhysRevB.82.113406http://dx.doi.org/10.1007/s00214-013-1350-zhttp://dx.doi.org/10.1007/s00214-013-1350-zhttp://dx.doi.org/10.1007/s00214-013-1350-zhttp://dx.doi.org/10.1007/s00214-013-1350-zhttp://dx.doi.org/10.1021/jp207571nhttp://dx.doi.org/10.1021/jp207571nhttp://dx.doi.org/10.1021/jp207571nhttp://dx.doi.org/10.1021/jp207571nhttp://dx.doi.org/10.1103/PhysRevB.86.155151http://dx.doi.org/10.1103/PhysRevB.86.155151http://dx.doi.org/10.1103/PhysRevB.86.155151http://dx.doi.org/10.1103/PhysRevB.86.155151http://dx.doi.org/10.1103/PhysRevB.78.045409http://dx.doi.org/10.1103/PhysRevB.78.045409http://dx.doi.org/10.1103/PhysRevB.78.045409http://dx.doi.org/10.1103/PhysRevB.78.045409http://dx.doi.org/10.1021/ja3117735http://dx.doi.org/10.1021/ja3117735http://dx.doi.org/10.1021/ja3117735http://dx.doi.org/10.1021/ja3117735http://dx.doi.org/10.1063/1.2432410http://dx.doi.org/10.1063/1.2432410http://dx.doi.org/10.1063/1.2432410http://dx.doi.org/10.1063/1.2432410http://www.spectrafox.comhttp://dx.doi.org/10.1016/j.cpc.2004.12.014http://dx.doi.org/10.1016/j.cpc.2004.12.014http://dx.doi.org/10.1016/j.cpc.2004.12.014http://dx.doi.org/10.1016/j.cpc.2004.12.014http://dx.doi.org/10.1063/1.2770708http://dx.doi.org/10.1063/1.2770708http://dx.doi.org/10.1063/1.2770708http://dx.doi.org/10.1063/1.2770708http://dx.doi.org/10.1103/PhysRevLett.80.890http://dx.doi.org/10.1103/PhysRevLett.80.890http://dx.doi.org/10.1103/PhysRevLett.80.890http://dx.doi.org/10.1103/PhysRevLett.80.890http://dx.doi.org/10.1063/1.3382344http://dx.doi.org/10.1063/1.3382344http://dx.doi.org/10.1063/1.3382344http://dx.doi.org/10.1063/1.3382344http://dx.doi.org/10.1103/PhysRevB.31.805http://dx.doi.org/10.1103/PhysRevB.31.805http://dx.doi.org/10.1103/PhysRevB.31.805http://dx.doi.org/10.1103/PhysRevB.31.805http://dx.doi.org/10.1021/nl0728874http://dx.doi.org/10.1021/nl0728874http://dx.doi.org/10.1021/nl0728874http://dx.doi.org/10.1021/nl0728874http://dx.doi.org/10.1038/nmat2166http://dx.doi.org/10.1038/nmat2166http://dx.doi.org/10.1038/nmat2166http://dx.doi.org/10.1038/nmat2166http://dx.doi.org/10.1107/S056773946800094Xhttp://dx.doi.org/10.1107/S056773946800094Xhttp://dx.doi.org/10.1107/S056773946800094Xhttp://dx.doi.org/10.1107/S056773946800094Xhttp://dx.doi.org/10.1107/S0365110X52001064http://dx.doi.org/10.1107/S0365110X52001064http://dx.doi.org/10.1107/S0365110X52001064http://dx.doi.org/10.1107/S0365110X52001064http://dx.doi.org/10.1088/1367-2630/10/4/043033http://dx.doi.org/10.1088/1367-2630/10/4/043033http://dx.doi.org/10.1088/1367-2630/10/4/043033http://dx.doi.org/10.1088/1367-2630/10/4/043033http://dx.doi.org/10.1103/PhysRevB.86.235439http://dx.doi.org/10.1103/PhysRevB.86.235439http://dx.doi.org/10.1103/PhysRevB.86.235439http://dx.doi.org/10.1103/PhysRevB.86.235439http://dx.doi.org/10.1021/nl203530thttp://dx.doi.org/10.1021/nl203530thttp://dx.doi.org/10.1021/nl203530thttp://dx.doi.org/10.1021/nl203530thttp://dx.doi.org/10.1103/PhysRevB.88.201406http://dx.doi.org/10.1103/PhysRevB.88.201406http://dx.doi.org/10.1103/PhysRevB.88.201406http://dx.doi.org/10.1103/PhysRevB.88.201406http://dx.doi.org/10.1103/PhysRevLett.108.066804http://dx.doi.org/10.1103/PhysRevLett.108.066804http://dx.doi.org/10.1103/PhysRevLett.108.066804http://dx.doi.org/10.1103/PhysRevLett.108.066804http://dx.doi.org/10.1103/PhysRevLett.108.206805http://dx.doi.org/10.1103/PhysRevLett.108.206805http://dx.doi.org/10.1103/PhysRevLett.108.206805http://dx.doi.org/10.1103/PhysRevLett.108.206805http://dx.doi.org/10.1103/PhysRevLett.40.1525http://dx.doi.org/10.1103/PhysRevLett.40.1525http://dx.doi.org/10.1103/PhysRevLett.40.1525http://dx.doi.org/10.1103/PhysRevLett.40.1525http://dx.doi.org/10.1103/PhysRevB.31.4911http://dx.doi.org/10.1103/PhysRevB.31.4911http://dx.doi.org/10.1103/PhysRevB.31.4911http://dx.doi.org/10.1103/PhysRevB.31.4911http://dx.doi.org/10.1103/PhysRevB.62.R16341http://dx.doi.org/10.1103/PhysRevB.62.R16341http://dx.doi.org/10.1103/PhysRevB.62.R16341http://dx.doi.org/10.1103/PhysRevB.62.R16341http://dx.doi.org/10.1016/S0039-6028(01)00704-Xhttp://dx.doi.org/10.1016/S0039-6028(01)00704-Xhttp://dx.doi.org/10.1016/S0039-6028(01)00704-Xhttp://dx.doi.org/10.1016/S0039-6028(01)00704-Xhttp://dx.doi.org/10.1016/S0039-6028(03)00151-1http://dx.doi.org/10.1016/S0039-6028(03)00151-1http://dx.doi.org/10.1016/S0039-6028(03)00151-1http://dx.doi.org/10.1016/S0039-6028(03)00151-1http://dx.doi.org/10.1103/PhysRevB.24.2951http://dx.doi.org/10.1103/PhysRevB.24.2951http://dx.doi.org/10.1103/PhysRevB.24.2951http://dx.doi.org/10.1103/PhysRevB.24.2951http://dx.doi.org/10.1016/j.orgel.2006.03.010http://dx.doi.org/10.1016/j.orgel.2006.03.010http://dx.doi.org/10.1016/j.orgel.2006.03.010http://dx.doi.org/10.1016/j.orgel.2006.03.010http://dx.doi.org/10.1103/PhysRevB.28.465http://dx.doi.org/10.1103/PhysRevB.28.465http://dx.doi.org/10.1103/PhysRevB.28.465http://dx.doi.org/10.1103/PhysRevB.28.465http://dx.doi.org/10.1103/PhysRevB.61.2263http://dx.doi.org/10.1103/PhysRevB.61.2263http://dx.doi.org/10.1103/PhysRevB.61.2263http://dx.doi.org/10.1103/PhysRevB.61.2263http://dx.doi.org/10.1103/PhysRevLett.100.156805http://dx.doi.org/10.1103/PhysRevLett.100.156805http://dx.doi.org/10.1103/PhysRevLett.100.156805http://dx.doi.org/10.1103/PhysRevLett.100.156805http://dx.doi.org/10.1126/science.288.5470.1399http://dx.doi.org/10.1126/science.288.5470.1399http://dx.doi.org/10.1126/science.288.5470.1399http://dx.doi.org/10.1126/science.288.5470.1399http://dx.doi.org/10.1103/PhysRevLett.55.991http://dx.doi.org/10.1103/PhysRevLett.55.991http://dx.doi.org/10.1103/PhysRevLett.55.991http://dx.doi.org/10.1103/PhysRevLett.55.991http://dx.doi.org/10.1103/PhysRevLett.55.987http://dx.doi.org/10.1103/PhysRevLett.55.987http://dx.doi.org/10.1103/PhysRevLett.55.987http://dx.doi.org/10.1103/PhysRevLett.55.987http://dx.doi.org/10.1016/0038-1101(66)90071-2http://dx.doi.org/10.1016/0038-1101(66)90071-2http://dx.doi.org/10.1016/0038-1101(66)90071-2http://dx.doi.org/10.1016/0038-1101(66)90071-2http://dx.doi.org/10.1098/rspa.1928.0091http://dx.doi.org/10.1098/rspa.1928.0091http://dx.doi.org/10.1098/rspa.1928.0091http://dx.doi.org/10.1098/rspa.1928.0091http://dx.doi.org/10.1016/S0921-4526(99)02884-7http://dx.doi.org/10.1016/S0921-4526(99)02884-7http://dx.doi.org/10.1016/S0921-4526(99)02884-7http://dx.doi.org/10.1016/S0921-4526(99)02884-7http://dx.doi.org/10.1103/PhysRevB.71.241404http://dx.doi.org/10.1103/PhysRevB.71.241404http://dx.doi.org/10.1103/PhysRevB.71.241404http://dx.doi.org/10.1103/PhysRevB.71.241404http://dx.doi.org/10.1021/jp901226qhttp://dx.doi.org/10.1021/jp901226qhttp://dx.doi.org/10.1021/jp901226qhttp://dx.doi.org/10.1021/jp901226qhttp://dx.doi.org/10.1103/PhysRevB.72.115404http://dx.doi.org/10.1103/PhysRevB.72.115404http://dx.doi.org/10.1103/PhysRevB.72.115404http://dx.doi.org/10.1103/PhysRevB.72.115404http://dx.doi.org/10.1103/PhysRevB.76.195404http://dx.doi.org/10.1103/PhysRevB.76.195404http://dx.doi.org/10.1103/PhysRevB.76.195404http://dx.doi.org/10.1103/PhysRevB.76.195404http://dx.doi.org/10.1063/1.2825595http://dx.doi.org/10.1063/1.2825595http://dx.doi.org/10.1063/1.2825595http://dx.doi.org/10.1063/1.2825595

FABIAN SCHULZ et al. PHYSICAL REVIEW B 89, 235429 (2014)

[63] B. Wang, M. Caffio, C. Bromley, H. Fruechtl, and R. Schaub,ACS Nano 4, 5773 (2010).

[64] H. B. Michaelson, J. Appl. Phys. 48, 4729 (1977).[65] C. L. Lin, S. M. Lu, W. B. Su, H. T. Shih, B. F. Wu, Y. D. Yao,

C. S. Chang, and T. T. Tsong, Phys. Rev. Lett. 99, 216103 (2007).[66] L. Vitali, G. Levita, R. Ohmann, A. Comisso, A. De Vita, and

K. Kern, Nat. Mater. 9, 320 (2010).

[67] H. Ishii, K. Sugiyama, E. Ito, and K. Seki, Adv. Mater. 11, 605(1999).

[68] J. Hwang, A. Wan, and A. Kahn, Mater. Sci. Eng. R-Rep. 64, 1(2009).

[69] R. Smoluchowski, Phys. Rev. 60, 661 (1941).[70] N. D. Lang and W. Kohn, Phys. Rev. B 3, 1215 (1971).[71] R. M. Nieminen and C. H. Hodges, J. Phys. F 6, 573 (1976).

235429-8

http://dx.doi.org/10.1021/nn101520khttp://dx.doi.org/10.1021/nn101520khttp://dx.doi.org/10.1021/nn101520khttp://dx.doi.org/10.1021/nn101520khttp://dx.doi.org/10.1063/1.323539http://dx.doi.org/10.1063/1.323539http://dx.doi.org/10.1063/1.323539http://dx.doi.org/10.1063/1.323539http://dx.doi.org/10.1103/PhysRevLett.99.216103http://dx.doi.org/10.1103/PhysRevLett.99.216103http://dx.doi.org/10.1103/PhysRevLett.99.216103http://dx.doi.org/10.1103/PhysRevLett.99.216103http://dx.doi.org/10.1038/nmat2625http://dx.doi.org/10.1038/nmat2625http://dx.doi.org/10.1038/nmat2625http://dx.doi.org/10.1038/nmat2625http://dx.doi.org/10.1002/(SICI)1521-4095(199906)11:83.0.CO;2-Qhttp://dx.doi.org/10.1002/(SICI)1521-4095(199906)11:83.0.CO;2-Qhttp://dx.doi.org/10.1002/(SICI)1521-4095(199906)11:83.0.CO;2-Qhttp://dx.doi.org/10.1002/(SICI)1521-4095(199906)11:83.0.CO;2-Qhttp://dx.doi.org/10.1016/j.mser.2008.12.001http://dx.doi.org/10.1016/j.mser.2008.12.001http://dx.doi.org/10.1016/j.mser.2008.12.001http://dx.doi.org/10.1016/j.mser.2008.12.001http://dx.doi.org/10.1103/PhysRev.60.661http://dx.doi.org/10.1103/PhysRev.60.661http://dx.doi.org/10.1103/PhysRev.60.661http://dx.doi.org/10.1103/PhysRev.60.661http://dx.doi.org/10.1103/PhysRevB.3.1215http://dx.doi.org/10.1103/PhysRevB.3.1215http://dx.doi.org/10.1103/PhysRevB.3.1215http://dx.doi.org/10.1103/PhysRevB.3.1215http://dx.doi.org/10.1088/0305-4608/6/4/017http://dx.doi.org/10.1088/0305-4608/6/4/017http://dx.doi.org/10.1088/0305-4608/6/4/017http://dx.doi.org/10.1088/0305-4608/6/4/017